Method and system for monitoring a joining process
The method improves the monitoring of joining processes by evaluating multiple features of pressing force data to assess the quality of composite product formation, reducing defects and improving process reliability in forming composite products like blister packs.
Patent Information
- Application Number
- EP2024167103
- Authority / Receiving Office
- EP · EP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-06-12
- Filing Date
- 2024-03-27
- Publication Date
- 2024-11-27
AI Technical Summary
Existing joining processes for forming composite products, such as blister packs, face quality issues due to foreign bodies between substrates, which current monitoring methods, relying on threshold comparisons of pressing force, cannot detect with absolute certainty, leading to potential defects in the final product.
A computer-implemented method for automatically monitoring the joining process by acquiring and evaluating measurement data to extract features representing the course of pressing force, determining an evaluation indicator, and controlling the pressing device based on the evaluation result, using a combination of characteristic features to assess the quality of the joining process and prevent defects.
This method enhances the reliability and accuracy of joining process monitoring, reducing the number of mistakenly classified defective products and making the process more robust against variations, such as thermal expansion, by using multiple characteristic features rather than single measurable parameters.
Smart Images

Figure IMGAF001_ABST
Abstract
Description
[0001] The present invention relates to a method, in particular a computer-implemented method, for automatically monitoring a joining process for joining a plurality of overlapping, in particular stacked, substrates to one another to form a composite product under pressure from a pressing device. The joining process can specifically be a sealing process for producing packaging, such as blister packs. Accordingly, the pressing device can in particular be a sealing machine, such as a plate sealing machine. The invention further relates to a data processing system configured to carry out the method and a computer program likewise configured to carry out the method, as well as a system comprising the data processing system for carrying out and automatically monitoring a joining process.
[0002] In joining processes of the type mentioned above, there is a risk that the quality of the resulting composite products will be impaired if there is a foreign body between the substrates to be joined during the joining process at a point where the substrates are to be brought into contact for their connection, or if there is a product to be enclosed between the substrates, e.g. for the purpose of its packaging.
[0003] Such a case can arise particularly in the production of blister packs, for example for food, pharmaceutical or cosmetic products. Such blister packs often have a first substrate which is in the form of a film with recesses formed therein for receiving the product to be packaged, e.g. one tablet per recess. A second, also usually film-like, substrate is applied to the first substrate by the joining process and connected under pressure by a pressing device, in particular in a form-fitting or material-fitting manner, such that the two substrates are connected outside the areas occupied by the recesses, while the product to be packaged lies in the recesses and is surrounded on all sides by the interconnected substrates.Such a joining process is often referred to as a "sealing process," and a device configured to carry it out is referred to as a "sealing machine" or "sealing device." The above-mentioned quality issues arise particularly when the product to be packaged, such as one or more tablets, is not properly positioned within its respective pockets during the pressurization process, but is located entirely or partially on substrate surfaces located between the pockets.
[0004] Since the occurrence of such quality problems cannot usually be ruled out with absolute certainty, the joining processes are usually monitored in order to detect such problems.
[0005] For this purpose, it is particularly known to measure the pressing force occurring on the pressing device during the pressurization of the substrates to be joined during the joining process and to perform a threshold comparison in this regard. If the pressing force is above a predetermined threshold, a defect in the joining process is concluded; otherwise, the joining process and the resulting joined product are considered defect-free.
[0006] It is an object of the invention to further improve the monitoring of joining processes of the type mentioned above.
[0007] This problem is solved according to the teaching of the independent claims. Various embodiments and further developments of the solution are the subject of the dependent claims.
[0008] A first aspect of the solution presented here relates to a method, particularly a computer-implemented method, for automatically monitoring a joining process for connecting, particularly laminating, a plurality of overlapping substrates to form a composite product under pressure applied by a pressing device, particularly a sealing machine, e.g., a plate sealing machine. The method comprises: (i) Acquiring, in particular receiving, reading in or measuring data which represent respective measured values of a, in particular scalar or vectorial, measured variable for different working positions assumed by the pressing device during the application of pressure, which indicates a pressing force acting on the substrates during the respective working position or a variable dependent thereon; (i) Evaluating the acquired measurement data in order to extract therefrom, in particular to calculate, a plurality of different features which each individually and / or jointly represent at least one characteristic property of a profile of the measured variable represented by the measurement data as a function of the working positions of the pressing device or of a variable corresponding to the working positions;(iii) determining an evaluation indicator, in particular a numerical one, defined as a measure of the quality of the joining process depending on the extracted features; (iv) evaluating the joining process based on a comparison, for example in the form of a comparison, of the determined evaluation indicator with a defined, in particular numerical, evaluation reference, thereby generating an evaluation result, in particular a numerical one, resulting from the comparison (as a measure of the quality of the joining process or of the composite product produced thereby); and (v) controlling the pressing device and / or further processing, in particular further processing, ejecting or marking, of the composite product produced by means of the joining process depending on the evaluation result.
[0009] The term "substrate," as used herein, refers to a homogeneous or heterogeneous, particularly film- or plate-shaped solid (i.e., having a solid state of aggregation at standard temperature of 20°C and standard pressure of 1013 hPa), which is suitable for being combined with one or more other such or similar solids under pressure to form a composite product. In particular, a substrate can be formed as a plastic or metal foil, or can itself already be a composite material. In particular, the various layers of a blister pack each fall under the term "substrate," as used herein.
[0010] The term "characteristic feature" or "feature" for short, as used herein, refers to a property of the measured variable profile that is generally suitable for distinguishing measured variable profiles that occur during a flawed joining process from other measured variable profiles that occur during a flawed joining process. Such flawed executions can particularly affect cases in which the quality of the resulting composite products is impaired by the presence of one or more foreign bodies or products to be packaged between the substrates to be joined during the joining process at a location where the substrates are to be brought into contact for their connection. While a feature can optionally have a direct physical meaning, it can also be merely a calculated value without any direct significance.What is crucial, however, is that it can be used alone or through its contribution in combination with other features to detect an anomaly occurring during the joining process.
[0011] The term "quantity," as used herein, refers to a physical or chemical quantity, i.e., a quantitatively determinable property of a process or state in a physical or chemical object. In particular, pressures, forces, electrical or mechanical stresses, material compositions or structures, mass densities, lengths, etc., are all quantities in this sense.
[0012] The term "working position" as used herein is to be understood as a respective state of the pressing device, wherein different working positions correspond to different working points, in particular strokes or distances of pressing tools of the pressing device which exert a pressing effect on the material to be pressed (in this case substrates) during pressing.
[0013] The terms "comprises," "includes," "includes," "has," "has," "with," or any other variation thereof, as used herein, are intended to cover non-exclusive inclusion. For example, a method or apparatus that includes or has a list of elements is not necessarily limited to those elements, but may include other elements not expressly listed or that are inherent in such a method or apparatus.
[0014] Furthermore, unless explicitly stated otherwise, "or" refers to an inclusive "or" rather than an exclusive "or." For example, a condition A or B is satisfied by one of the following conditions: A is true (or present) and B is false (or absent), A is false (or absent) and B is true (or present), and both A and B are true (or present).
[0015] The terms "a" or "an" as used herein are defined as "one or more." The terms "another" and "another," and any other variations thereof, are defined as "at least one other."
[0016] The term "plurality" as used herein shall mean "two or more".
[0017] The terms "first," "second," "third," and similar terms in the description and claims are used to distinguish between similar or otherwise identically named elements and not necessarily to describe a sequential, spatial, or chronological order. It is understood that the terms so used are interchangeable under appropriate circumstances, and that the embodiments of the solution described herein may also function in orders other than those described or illustrated herein.
[0018] The term "configured" or "set up" to perform a specific function (and respective variations thereof), as used here, is to be understood that a relevant device or component thereof is already in a configuration or setting in which it can perform the function, or is at least adjustable—i.e., configurable—so that it can perform the function after being set accordingly. The configuration can be achieved, for example, by setting parameters of a process sequence or switches or the like to activate or deactivate functionalities or settings. In particular, the device can have several predetermined configurations or operating modes, so that configuration can be achieved by selecting one of these configurations or operating modes.
[0019] The method according to the first aspect is particularly characterized by the fact that a number of different characteristic features of the measured variable's profile are used to evaluate the joining process. This has the advantage that not only directly measurable features but also suitable features that can only be derived indirectly from the profile can be used for the evaluation, thus expanding the available evaluation basis.
[0020] Using multiple characteristics instead of a single characteristic as the basis for evaluation allows for an increase in achievable reliability (or, equivalently, a reduction in the error rate) and / or the accuracy of the evaluation. This allows for better optimization of production and increased yield of the joining process while reducing the number of composite products mistakenly classified as defective. In particular, the method is therefore not limited to the detection of very large overload situations, unlike, for example, a simple measurement of a maximum force followed by a threshold comparison. In particular, the method is also less susceptible to any thermal expansion of components of the press device or the pressed material that could distort measurements.
[0021] In the following, various exemplary embodiments of the method are first described, which can each be combined with each other as well as with the other aspects of the present solution described below, unless this is expressly excluded or is technically impossible.
[0022] In some embodiments, at least one of the features is or will be selected from the following list of variables or defined as a variable that is in a dependent relationship with at least one variable from the list: an extreme value corresponding to a maximum pressing force, in particular a local or global maximum, of the course of the measured variable; an indicator which characterises a working position at which the extreme value occurred; a pre-tensioning force, in particular a constant one, to which the arrangement of the overlapping substrates was exposed by the pressing device before the pressure application carried out to join the substrates with a pressing force which was increased in comparison, in particular abruptly; a post-tensioning force, in particular a constant one, to which the arrangement of the then already connected substrates was exposed by the pressing device after the pressure application carried out to join the substrates with a pressing force which was increased in comparison, in particular abruptly; a difference or a ratio of pre-tensioning force and post-tensioning force;an indicator which characterises a working position in which, starting from the pre-tensioning force, the pressure application for joining the substrates, carried out with a comparatively, in particular abruptly, increased pressing force, began or increased, an indicator which characterises a working position in which, after the pressure application for joining the substrates, carried out with a comparatively, in particular abruptly, increased pressing force, the post-tensioning force began or the preceding pressure application decreased, and exceeded a second predetermined force threshold; an integral of the measured variable determined over its course.
[0023] A (first) quantity is "dependent" on at least one (second) quantity from the list if it can be determined in a predetermined manner, in particular in the sense of a (particularly injective) mathematical function, from the at least one (second) quantity from the list on which it depends. In particular, such a dependency relationship can also be defined using mathematical differentiation or integration.
[0024] The pre-tensioning and / or post-tensioning can be achieved in particular by means of a spring force acting on the substrates or the composite product or by means of gas pressure, for example to secure the as yet unbonded substrates against slipping relative to one another (in the case of pre-tensioning) or, in both cases, to secure the substrates or the composite product produced therefrom from slipping relative to a support. The means used to effect the pre-tensioning and / or post-tensioning can differ in particular from the means used for pressure application. For example, the pre-tensioning and / or post-tensioning can be achieved solely by a spring force, while the pressure application is achieved by a pressure piston or a motor-driven actuator.
[0025] The characteristics listed in the list (and, where applicable, the variables that are dependent on them) have proven to be particularly suitable variables for characterising faulty executions compared to faultless executions of the joining process using a combination of two or more of these variables.
[0026] In some embodiments, the measurement data acquired is sensor data that has been or will be acquired based on measurements with at least one of the following sensor types: load cell, strain gauge, accelerometer. These sensor types have in common that they directly or indirectly enable a measurement of the pressing force exerted by the pressing device on the material to be pressed, depending on its respective working position.
[0027] In some embodiments, the measurement data is subjected to preprocessing to improve data quality, and at least one of the features is extracted from the measurement data prepared during the preprocessing. For this purpose, the preprocessing may, in particular, include filtering the measurement data to reduce or eliminate high-frequency interference. This can increase the quality of the subsequent feature extraction and, consequently, reduce the likelihood of resulting misinterpretations.
[0028] In some embodiments, the preprocessing comprises transforming the measurement data such that the transformed measurement data continue to represent the measured variable as a function of the working positions of the pressing device or the variable corresponding to the working positions, but are independent of the speed of the pressure application exerted by the pressing device (i.e., of the mathematical derivative of the pressing pressure or the pressing force with respect to time). The result of the preprocessing thus represents the measurement data, such as a force curve, with reference to a speed-independent variable, such as a position of a pressing tool of the pressing device that characterizes a working position. As a result, the speed plays no role in the subsequent determination of the features.Since small speed fluctuations are often unavoidable, the transformation serves in particular to make the process even more robust against disturbances or against process fluctuations that cannot be controlled or are difficult to control.
[0029] In some embodiments, the evaluation indicator is determined depending on the extracted features by weighting the features with respective assigned weight values. This can be done, in particular, by forming a weighted sum. The weighting of the features can be defined, in particular, based on respective weight factors assigned individually to the features. This makes it possible to incorporate the various features used to monitor the joining process to varying degrees into the determination of the evaluation indicator. If the weights can also be determined variably, the evaluation indicator can be achieved, especially iteratively, as a very good approximation of an "ideal" evaluation indicator.
[0030] In some such embodiments, the features are each quantified based on at least one respective value from a predefined, limited set of values that is individual for each feature or the same for all features. These values, along with the respective assigned weight values, are incorporated into the determination of the evaluation indicator. For example, the set of values can be defined by the numerical range [0; 1]. Such standardization facilitates the implementation and efficiency of the method. In particular, it allows for equal or only slightly differing accuracies to be achieved with regard to the quantification of the individual features.
[0031] In some embodiments, at least one of the weight values is derived based on the respective evaluation results of one or more previous executions of the joining process, wherein these evaluation results each indicate a faulty joining process. Determining the weight values based on faulty process executions allows for a particularly effective and rapid convergence of the weight values toward a respective ideal value. Several process executions with different error causes are used as a basis in order to be able to correctly detect a multitude of different errors later after the weight values have been determined.
[0032] In some embodiments, the evaluation indicator is determined using a trained machine learning model with the extracted features as input variables of the machine learning model. In particular, an artificial neural network (ANN) is considered as the machine learning model, whose parameters (weights of its neurons) are optimized during training with suitable training data. The training data can, in particular, comprise a plurality of different data sets, each of which contains a possible value assignment for the features serving as input variables as well as a correct value for the evaluation indicator to be assigned to it.Compared to a once-defined weighting scheme, the use of a machine learning model has the particular advantage that further continuous optimization of the process is available even after its initial commissioning and that even very complex weighting functions that would otherwise be difficult to determine can be determined or approximated very well.
[0033] In some embodiments, the comparison of the determined evaluation indicator with a defined evaluation reference comprises a comparison of the determined evaluation indicator with at least one predetermined evaluation threshold serving as an evaluation reference. The evaluation indicator, or the rule, e.g., a weighting function, for determining it, can in particular be defined such that the set of values for the evaluation indicator corresponds to a specific, in particular bilaterally limited, value range (e.g., [0; 1]), so that the evaluation threshold can be determined in a simplified manner only with reference to this value range.
[0034] In some embodiments, the method comprises monitoring multiple, in particular sequential or parallel, executions of the joining process to produce a corresponding number of composite products. Furthermore, the evaluation is performed for each monitored joining process. For each joining process, the respective evaluation result is stored in a data structure, in particular in a shift register, or database in such a way that the evaluation results are assigned to the respective joining processes and the respective individual evaluation results of each of the joining processes can be inferred from the contents stored in the data structure or database. Thus, the method can be extended to a plurality, in particular a multiplicity, of process executions in such a way that the respective evaluation results, in particular several at the same time, remain at least temporarily retrievable even after the respective execution of the joining process.In particular, in the case of a production line, the evaluation results can thus "run along with the composite products" along the further production process, so that they also remain evaluable for any work stations located downstream of the pressing device in the process flow, e.g. a station for the removal of defective parts.
[0035] In some embodiments, the pressing device is controlled for each joining process depending on the respective assigned evaluation result stored in the data structure or database. In particular, it is conceivable that the pressing process is stopped or executed in a different manner than in the error-free case if an error is detected during process execution based on the stored evaluation result. This can be used, in particular, to prevent or reduce any avoidable damage or contamination to the pressing device, for example, due to crushed foreign bodies or objects to be packaged.
[0036] In some embodiments, the evaluation results are transmitted to a controller of the pressing device using communication based on the known OPC Unified Architecture (OPC-UA) information model, in particular according to its version 2 (e.g., version 2.0), to enable and / or prompt the controller to control the pressing device depending on the evaluation result. Thus, the evaluation results can be transmitted in a standardized and thus manufacturer-independent and reliable manner.
[0037] In some embodiments, a sealing machine, in particular a plate sealing machine, is used as the pressing device to combine two or more of the substrates to form a blister pack. Such a blister pack can be configured in particular for packaging tablet-shaped pharmaceuticals, dietary supplements, or cosmetic products. Especially with blister packs, there is a fundamental risk that the objects to be packaged (e.g., tablets) will end up outside the pockets provided for their reception, thereby disrupting the joining process and leading to defects in the resulting composite product. Furthermore, with such products, the requirements for the integrity of the products and their packaging are usually particularly high, so high-quality monitoring of their production and packaging is particularly relevant here.
[0038] A second aspect of the present solution relates to a data processing system for automatically monitoring a joining process, wherein the data processing system has at least one processor platform (with at least one processor) and a memory coupled thereto, in which instructions are stored which, when executed on the processor platform, cause it to carry out the method according to the first aspect, in particular according to one or more of its embodiments described herein.
[0039] A third aspect of the present solution relates to a system for executing and automatically monitoring a joining process for joining a plurality of overlapping substrates to one another to form a composite product under pressure from a pressing device, in particular a plate sealing machine. The system comprises the pressing device, including a controller therefor, and the data processing system according to the second aspect, which is configured to communicate evaluation results for the joining process determined by it according to the method (i.e., according to the method according to the first aspect) to the controller. The data processing system and the controller can, in particular, also coincide, so that the data processing system is then simultaneously configured to assume the functions of the controller.
[0040] A fourth aspect of the present solution relates to a computer program or computer program product, in particular a non-volatile computer-readable storage medium, each with instructions which, when executed on a data processing system according to the second aspect, cause the system to carry out the method according to the first aspect.
[0041] The computer program (or the instructions) can be stored, in particular, on a non-volatile data storage medium. This is preferably a data storage medium of the system according to the third aspect, for example a data storage medium in the form of an optical or magnetic data storage medium or a flash memory module or another non-volatile semiconductor memory. In another implementation, the computer program can be present as a file on a system-external data processing unit, in particular on a server, and can be downloaded to the data processing system according to the second aspect via a data connection, for example the Internet or a dedicated data connection, such as a proprietary or local network. In addition, the computer program can have a plurality of interacting individual program modules. The modules can, in particular, be configured, or at least be usable, to function in the sense of distributed computing (DC)."Distributed computing" is carried out on different devices (computers or processor units) that are geographically separated from each other and connected via a data network.
[0042] The features and benefits explained with regard to the first aspect of the solution also apply to the other aspects of the solution.
[0043] The method according to the first aspect is preferably designed for implementation using the data processing system according to the second aspect, in particular an embodiment thereof described herein. The data processing system, in turn, is preferably designed to implement the method according to the first aspect, in particular an embodiment thereof described herein.
[0044] Further advantages, features and possible applications of the present solution will become apparent from the following detailed description in the context of the figures.
[0045] It shows: Fig. 1 a flow chart illustrating an exemplary embodiment of the method for automatically monitoring a joining process. Fig. 2A an example of a temporal progression of a recorded measured value; Fig. 2B an exemplary temporal progression of a position coordinate, the values of which correspond to different working positions of a pressing device used according to the method; Fig. 2C schematically a preprocessing of the measured variable from Fig. 2A and the combination of its course with the course of the position coordinate from Fig. 2B to obtain a curve of the preprocessed measured value as a function of the position coordinate; Fig. 3A a scheme to illustrate an extraction of various features from the course of the measured variable, in particular according to Fig. 2C ; Fig. 3B-D various diagrams to illustrate feature extraction for three different exemplary features; Fig. 4A a flow chart illustrating a process for determining a quantified rating assigned to a particular characteristic, here using the example of the characteristic "maximum force"; Fig. 4B a numerical example of the process Fig. 4A ; Fig. 5A a flow chart illustrating a process for determining an evaluation indicator for the joining process using a weighted sum of the respective evaluations of the individual characteristics; Fig. 5B a numerical example of the process Fig. 5A ; and Fig. 6 schematically an exemplary embodiment of a system for automatically monitoring a joining process, designed to carry out the method from the preceding figures.
[0046] In the figures, like reference numerals designate like, similar, or corresponding elements. Elements shown in the figures are not necessarily drawn to scale. Rather, the various elements shown in the figures are depicted in such a way that their function and general purpose will be understood by those skilled in the art. Connections and couplings between functional units and elements shown in the figures can, unless expressly stated otherwise, also be implemented as an indirect connection or coupling. Unless stated otherwise in detail or technically only possible in a specific way, functional units can, in particular, be implemented optionally as hardware, software, or a combination of hardware and software.
[0047] In Fig. 1 An exemplary embodiment 100 of the method for automatically monitoring a joining process is shown as a flow chart. The method 100 can be implemented in particular by the Fig. 6 illustrated system 600. Therefore, the method 100 will be explained below with exemplary reference to the system 600, which will therefore be described first.
[0048] The system 600 comprises a pressing device 605 for carrying out a joining process in which two or more substrates, e.g. plastic or metal foils, are pressed together under pressure by means of a pressing force in order to connect the substrates in a material-locking or form-locking manner to form a composite product P. The pressing device 605 can in particular be, as in Fig. 6 Illustrated is a sealing machine, specifically a plate sealing machine, for producing blister packs. Such blister packs can be used, for example, for packaging pharmaceutical products, in particular tablets. Such a plate sealing machine can, in particular, as shown, have two pressing tools 105a and 105b, between which the substrates to be joined are inserted in an overlapping manner, in particular stacked on top of one another, in order to then be subjected to pressure or the pressing force between the pressing tools 105a,b.
[0049] The pressing device 605 has a control unit 610 for its control by means of suitable control signals C, which is designed in particular by means of a programmable logic controller (PLC).
[0050] Furthermore, the system 600 comprises a sensor system 615 for acquiring measurement data, which represents respective measured values of a measured variable for various working positions assumed by the pressing device 605 during the application of pressure. This measured variable indicates a pressing force exerted by the pressing tools 605a, b on the substrates during the respective working position or a variable dependent thereon, such as a pressing pressure or an acceleration corresponding to the pressing force. Suitable sensor types for this purpose include, in particular, load cells, strain gauges, and accelerometers.
[0051] The system 600 also includes a data processing system 620, such as a computer with a memory 625. The other Fig. 6 The components of the system 600 shown are introduced below in the explanation of the method 100.
[0052] Now, to explain the procedure 100, additional reference is made to Fig. 1 Reference is made to the method 100. The method 100 can be configured, in particular, as a computer-implemented method and implemented by means of a computer program that can be executed on the data processing system and stored, for example, in the memory 625.
[0053] Within the scope of the method 100, the measurement data for the at least one measured variable are first recorded in a process 105, for which purpose the measurement data are received by the sensor system 615.
[0054] An example of a temporal progression 200 of the measured value "pressing force" represented by recorded measurement data is shown in Fig. 2A illustrated. The time course covers a pressing process in which the substrates to be joined are first pre-tensioned against each other ("pre-tension", VSP), then connected with increased pressing force (pressing tension, PSP), and finally remain tensioned against each other again at a lower pressing force ("post-tension", NSP). The pre-tension can be achieved in particular (solely) by means of a spring force at an initial pressing force level that is not yet sufficient to connect the substrates, whereby the then not yet connected, overlapping substrates are clamped between the tools 605a, b to secure them against slipping, to eliminate any air pockets between the surface sections to be joined, and / or to eliminate unwanted deformations of the substrates. The connection under pressure with a higher pressing force can be carried out with a dedicated pressing mechanism, e.g.Hydraulically or using a motor to generate the pressing force. Like preload, the post-tensioning can be achieved solely or at least primarily by a spring force, although other designs are also conceivable.
[0055] The recorded measurement data also represent an example in Fig. 2B illustrated temporal progression 205 of the working position of the pressing device, in particular the relative position of the tools 605a and 605b to one another. In this case, the range of the working positions successively assumed during the execution of the joining process can be expediently standardized, for example, as in the present example, to a degree range [0°; 360°] for a complete process execution, which can, in particular, be periodically repeatable. The term "position" is used below to indicate the respective working position in relation to this degree range. Of course, other ways of quantifying different working positions are also possible instead.
[0056] After being recorded in process 105, the recorded measurement data are pre-processed in a further process 110, in particular as in Fig. 2C The preprocessing may in particular comprise filtering 215 (in particular low-pass filtering) of the temporal profile 200 of the measured variable(s) represented by the measurement data in order to remove any interference (in particular high-frequency interference, as in Fig. 2A recognizable) in it or at least to reduce it.
[0057] During preprocessing, the (optionally filtered) curve 200 and the curve 205 are also linked (typically mathematically) within the framework of a transformation 220 in order to obtain a curve of the measured variable(s) as a function of the working position (position). In the exemplary diagram 210 from Fig. 2C Two different curves of the measured variable as a function of position are already shown, whereby the curve with the lower maximum corresponds to an error-free process execution, while the other curve corresponds to an error-prone process execution.
[0058] In particular, imagine the case of a blister pack, in the production of which two substrates are to be joined together in such a way that the tablets to be packaged ideally only ever lie in one of the pockets formed on the substrates. The error-free curve in diagram 210 characterizes the joining process. Initially, only the essentially constant pre-tension is applied, before the force increases significantly at a position of approximately 80° and culminates in a maximum. Here, the two pressing tools 605a, b are pressed together to the maximum in order to join the substrates, while the tablets lying in the pockets are not subjected to the pressing force. During the subsequent pressure reduction, a step occurs in the force curve before the force then continuously decreases to the post-tension.
[0059] In the event of a defect, however, at least one of the tablets ends up at least partially outside its assigned zone, so that when the pressing tools 605a, b move together, an increase in the pressing force acting on the substrates is observed earlier (at approximately 70°) than in the defect-free process, because the aforementioned tablet(s) act as resistance. The maximum force is also correspondingly higher than in the defect-free process, and even the post-tensioning can begin later. The area or the integral under the curve profiles therefore also differs. These and possibly other differences in the profiles can be used below to carry out a quality assessment of the joining process.
[0060] An advantage of the representation according to Diagram 210 is that the measured variable curve is available regardless of the process speed, so that speed fluctuations during the further course of the process are not important. The analysis of the curves can thus be performed independently of the speed.
[0061] Within the framework of the method 100, a further process 115 follows, in which the pre-processed measurement data (cf. Fig. 2C ) various features characteristic of the course 210 are extracted. This is in Fig. 3A illustrated, where an example of a scheme 300 for the extraction of four different features in respective extraction processes 305 to 320 is shown.
[0062] Fig. 3B illustrates the extraction process 305 in more detail, in which a first feature "maximum force" is extracted, which characterizes the maximum pressing force occurring during the measurement. In Fig. 3B Here, for two different product types, each one is a type to be Fig. 2C The corresponding curve of the pressing force is shown, including the position of the respective maxima (extreme values) E1 and E2 occurring during the pressing phase following preloading. Their position can be calculated and analyzed using known methods, particularly using differential calculus. The maxima represent the desired values for the first characteristic.
[0063] Fig. 3C illustrates the extraction processes 310 and 315 in more detail, in which the second feature "threshold exceeded during the ascent" and the third feature "threshold exceeded during the descent" are extracted. These are the positions S1 and S2, respectively, where the force curve, starting from the pre-tensioning during the ascent, exceeds a defined increase force threshold F1 (second feature) or, before reaching the post-tensioning during the descent, exceeds a defined decrease force threshold F2. The distance between these two positions determined in this way is also suitable as a feature.
[0064] Fig. 3D illustrates the extraction process 320 in more detail, in which a fourth feature, "area integral," is extracted, which characterizes an integral under the curve of the pressing force, whereby (as shown) the position ranges of the prestress VSP and the poststress NSP can be disregarded in order to achieve a higher sensitivity of the feature. In Fig. 3D This is illustrated by way of example for two different curves, where the integrals A1 and A2 are determined as areas under the curves (in this example, area A1 also contains area A2 as a sub-area). This can be done in a well-known manner using methods of numerical integration.
[0065] Back on Fig. 1 Referring to this, in the method 100 a further process 120 follows, in which an evaluation indicator for the quality of the joining process is determined from the extracted features in several stages.
[0066] A first stage of this process 120 is in Fig. 4A and as a suitable concrete numerical example 401 in Fig. 4B This first stage involves performing an individual evaluation for each of the extracted features within a process 400. In the Figuren 4A und 4B Although this is done for the first characteristic "maximum force" (F max ) as an example, the same methodology can be applied to the other characteristics.
[0067] For this purpose, a repeated, in particular clocked, execution of the joining process is assumed, as is common in particular in the production of blister packs. First, in a current cycle n, an average value ØF max is calculated in step 410 from several, in this specific example four, measured values 405 of the immediately preceding cycles n-4 to n-1, i.e. the individual measured values F max,n-4 to F max,n-1 . In a further step 415, a deviation (here specifically a difference) X between this average value ØF max and a measured value F max,n taken from the measured data for the current cycle is determined. This deviation X is then compared in a step 420 with an evaluation threshold TH defined for the first feature. If the deviation is above this evaluation threshold TH 1, the feature is evaluated in a further step 435 with a first value (in this case with B 1 ="1"), otherwise with a different second value (e.g.B 1 = "0"). Thus, a binary value assignment is particularly possible here. All other attributes are evaluated accordingly, in particular using attribute-specific individual evaluation thresholds TH i , to obtain respective ratings B 2 to B 4 .
[0068] A second stage of process 120 is shown in the left part of a Fig. 5A process 500 and as a concrete numerical example 501 in Fig. 5B illustrated. In this second stage, the values B 1 to B 4 of the extracted features determined in the first stage are each weighted and combined with an individual weight value, in particular a respective weighting factor W 1 to W 4 , which can be done in particular by forming a sum of the weighted values (W i . B i ) of the features. The weighting factors W 1 to W 4 are expediently standardized according to the value range of the values B 1 to B 4 so that their sum results in the value 1. The result of the weighted sum is an evaluation indicator Y, the value of which in this case lies in the value interval [0; 1]. Optimal values for the weight values or in this case weighting factors, W 1 to W 4 can be determined in advance, in particular in the context of test series.Alternatively, the evaluation indicator Y can be determined using a trained machine learning model with the values B 1 to B 4 of the extracted features as input variables for the machine learning model, with the weighting factors W 1 to W 4 being determined automatically during training. An artificial neural network is particularly suitable as a machine learning model for this purpose.
[0069] On the basis of the evaluation indicator Y thus determined, an evaluation of the joining process can now be carried out in a further process 130 within the framework of the method 100 by comparing, in particular directly comparing, the evaluation indicator Y with a predetermined evaluation reference, in particular an evaluation threshold TH, wherein as a result of the comparison a corresponding evaluation result B is generated, which can be used as a quality evaluation for the joining process.
[0070] Finally, the pressing device 605, or more generally the system 600, and thus the further process sequence for treating the composite product P produced in the pressing device in the current cycle, can then be controlled as a function of the evaluation result B within the framework of a process 135. Such a control 135 can, in particular, also include a (re-)calibration of the pressing device. Another possibility, which can also be used cumulatively, is to separate good parts from defective parts depending on the evaluation results B for each cycle.
[0071] This is also in Fig. 6Illustrated, where in the system 600, a discharge device 630 is further provided downstream of the pressing device 605 in the process flow, which can be controlled by the data processing device 620 depending on the evaluation result B of the current cycle such that, if the evaluation result B indicates a faulty joining process, the produced composite product P is treated as a defective part and is discharged from the process by the discharge device 630 and transferred, for example, to a collection depot 635 for defective parts. Otherwise, if the evaluation result B indicates a fault-free joining process, the produced composite product P is treated as a good part and fed to the further process flow 640 for good parts.
[0072] In the case of the aforementioned repeated, particularly synchronized, execution of method 100, the evaluation results for each execution can be stored, in particular, in a data structure or database such that the evaluation results B are assigned to the respective executions of the joining process, so that the respective individual evaluation results of each of the joining processes can be inferred from the contents stored in the data structure or database. A shift register is particularly suitable as the data structure. The data structure or database can be stored, in particular, in the memory 625 of system 600 and configured to be accessible to the discharge device 630 and / or to additional workstations in system 600 used along the further process flow 640. In this way, the evaluation results can "migrate" through the process with the generated composite products P, i.e., remain continuously available.In particular, the communication of the evaluation results B within the system 600 can be carried out according to the OPC-UA information model (e.g. according to its version V. 2.0). LIST OF REFERENCE SYMBOLS
[0073] 100 Method for monitoring a joining process 105 - 135 Process steps of method 100 200 Time profile of the measured variable "pressing force" 205 Time profile of the working position of the pressing device 210 Profiles of the measured variable "pressing force" as a function of the working position 215 Preprocessing, in particular (low-pass) filtering 220 Transformation of profiles 200 (after preprocessing) and 205 to determine profile 210 300 Scheme for feature extraction 305 Extraction of a first feature "maximum force" 310 Extraction of a second feature "threshold exceeded during rise" 315 Extraction of a third feature "threshold exceeded during fall" 320 Extraction of a fourth feature "area integral" 400 Process for the individual evaluation of features,using the example of the first feature 401 numerical example of the process 400 405 measured values for the (first) feature for previous work cycles 410 averaging of the measured values 405 415 difference calculation 420 comparison with feature-related threshold 425 individual evaluation of the (first) feature 500 process for determining an evaluation indicator for the joining process 501 numerical example of the process 500 600 system for executing and automatically monitoring a joining process 605 pressing device, in particular sealing machine for blister packs 605a, b pressing tools 610 control unit of the pressing device, in particular PLC 615 sensors for recording measured values for measured variable(s) 620 data processing system 625 memory, in particular program memory and / or data memory 630 discharge device 635 collection depot for discharged defective parts 640further process flow for good parts BERvaluation indicator for joining process B i value of the i-th feature Ccontrol signals Ddata E1,E2 Maximum force values F max,j Measured values for the first characteristic "Maximum force" with j = n-4,..., n-1 ØF max Mean value of the measured values F max,j F max,n Current measured values for the first characteristic "Maximum force" F1 Increasing force threshold F2 Decreasing force threshold NSP Post-tensioning or post-tensioning force P Connection product PSP Press stress TH Evaluation threshold as evaluation reference for evaluation indicator TH i Individual evaluation threshold for the i-th characteristic VSP Pre-tensioning or pre-tensioning force W i Weight of the i-th characteristic X Deviation from the mean Y Weighted sum of the values of the characteristics,
Claims
1. A method (100) for automatically monitoring a joining process for joining a plurality of overlapping substrates to form a composite product under pressure application by a pressing device (605), wherein the method (100) comprises: acquiring (105) measurement data representing respective measured values of a measured variable for various working positions assumed by the pressing device (605) during the pressure application, which indicates a pressing force acting on the substrates during the respective working position or a variable dependent thereon; evaluating the acquired measurement data in order to extract a plurality of different features therefrom (115; 305,...,-320), each individually and / or jointly representing at least one characteristic property of a profile (210) of the measured variable represented by the measurement data as a function of the working positions of the pressing device (605) or a variable corresponding to the working positions; determining (120) an evaluation indicator (B) defined as a measure of the quality of the joining process as a function of the extracted features; evaluating (130) the joining process based on a comparison of the determined evaluation indicator (B) with a defined evaluation reference, generating an evaluation result resulting from the comparison; and controlling (135) the pressing device (605) and / or further processing, in particular further processing, discharging or marking, of the composite product produced by means of the joining process as a function of the evaluation result.
2. Method (100) according to claim 1, wherein at least one of the features is selected from the following list of variables or is defined as a variable that is dependent on at least one variable from the list: - a value that leads to a maximum pressing force (F max) corresponding extreme value (E1; E2) of the curve (210) of the measured variable; - an indicator which characterizes a working position in which the extreme value occurred; - a pre-tensioning force (VSP) to which the arrangement of the overlapping substrates was exposed by the pressing device (605) before the pressure application carried out to connect the substrates with a comparatively increased pressure force began; - a post-tensioning force (NSP) to which the arrangement of the then already connected substrates was exposed by the pressing device (605) after the pressure application carried out to connect the substrates with a comparatively increased pressure force had taken place; - a difference or a ratio of pre-tensioning force (VSP) and post-tensioning force (NSP);- an indicator which identifies a working position in which, starting from the pre-tensioning force (VSP), the pressure application to join the substrates, which was carried out with a comparatively increased pressing force, began or increased, and exceeded a first predetermined force threshold (F1); - an indicator which identifies a working position in which, after the pressure application to join the substrates, which was carried out with a comparatively increased pressing force, the post-tensioning force (NSP) began or the preceding pressure application decreased and exceeded a second predetermined force threshold (F2); - an integral of the measured variable determined over its course (210); 3. Method (100) according to one of the preceding claims, wherein the measurement data acquired are sensor data which were or are acquired on the basis of measurements with at least one of the following sensor types: load cell, strain gauge, accelerometer.
4. Method (100) according to one of the preceding claims, wherein the measurement data are subjected to preprocessing (110, 215) to improve the data quality, and the extraction of at least one of the features takes place from the measurement data prepared in the context of the preprocessing (110; 215).
5. The method (100) according to claim 4, wherein the preprocessing (110; 215) comprises filtering (215) the measurement data to reduce or eliminate high-frequency interference.
6. The method (100) according to claim 4 or 5, wherein the preprocessing (110; 215) comprises transforming the measurement data such that the transformed measurement data continue to represent the measurement variable as a function of the working positions of the pressing device (605) or the variable corresponding to the working positions, but are independent of the speed of the pressure application exerted by the pressing device (605).
7. Method (100) according to one of the preceding claims, wherein the determination of the evaluation indicator (B) as a function of the extracted features involves weighting the features with respective assigned weight values (W i ) takes place.
8. Method (100) according to claim 7, wherein the features are each quantified on the basis of at least one respective value from a predefined limited set of values that is individual for each feature or the same for all features, and these values are compared with the respectively assigned weight values (W i ) are included in the determination of the evaluation indicator (B).
9. The method (100) according to claim 7 or 8, wherein at least one of the weight values (W i ) is derived on the basis of the respective evaluation results of one or more previous executions of the joining process, whereby these evaluation results each indicate a faulty joining process.
10. Method (100) according to one of the preceding claims, wherein the determination of the evaluation indicator (B) is carried out using a trained machine learning model with the extracted features as input variables of the machine learning model 11. Method (100) according to one of the preceding claims, wherein the comparison of the determined evaluation indicator (B) with a defined evaluation reference comprises a comparison of the determined evaluation indicator (B) with at least one predetermined evaluation threshold serving as an evaluation reference.
12. The method (100) according to any one of the preceding claims, wherein: the method (100) comprises monitoring a multiple execution of the joining process to produce a corresponding number of composite products; and the evaluation is performed for each monitored joining process, wherein the respective evaluation result for each joining process is stored in a data structure or database such that the evaluation results are assigned to the respective joining processes therein, so that the respective individual evaluation results of each of the joining processes can be inferred from the contents stored in the data structure or database.
13. Method (100) according to one of the preceding claims, wherein the control of the pressing device (605) for each joining process takes place as a function of the respective assigned evaluation result stored in the data structure or database.
14. The method (100) according to any one of the preceding claims, wherein the evaluation results are transmitted to a controller (610) of the pressing device (605) using communication based on the OPC-UA information model in order to enable and / or cause it to control the pressing device (605) depending on the evaluation result.
15. The method (100) according to any one of the preceding claims, wherein a sealing machine is used as the pressing device (605) to join two or more of the substrates into a blister pack.
16. Data processing system (620) for automatically monitoring a joining process, wherein the data processing system (620) has at least one processor platform and a memory coupled thereto, in which instructions are stored which, when executed on the processor platform, cause it to carry out the method (100) according to one of the preceding claims.
17. System (600) for carrying out and automatically monitoring a joining process for joining a plurality of overlapping substrates to one another to form a composite product under pressure from a pressing device (605), wherein the system (600) comprises the pressing device (605) including a controller (610) therefor and the data processing system (620) according to claim 16, which is configured to communicate evaluation results for the joining process determined by it in accordance with the method to the controller (610).
18. Computer program or computer program product, in particular non-volatile computer-readable storage medium, each with instructions which, when executed on a data processing system (620) according to claim 16, cause the system to carry out the method (100) according to one of claims 1 to 15.
Citation Information
Patent Citations
Machine for forming. filling and sealing bags e.g. for potato crisps
DE19803838A1
Method and apparatus for sealing a web, and for making a packaging container
DE69822042T2
Joining device and joining method
EP3168032A1
Seal integrity monitoring and adaptive control method and apparatus
US5836136A
